Public Disclosure Authorized
Public Disclosure Authorized
Public Disclosure Authorized
Public Disclosure Authorized
WPS4503
P olicy R esearch W orking P aper
4503
Decentralization and the Provision of
Public Services:
Framework and Implementation
Aehyung Kim
The World Bank
Development Economics
Capacity Building, Partnership, and Outreach Team
January 2008
Policy Research Working Paper 4503
Abstract
This paper discusses decentralization (administrative,
fiscal and political) of government in public service
provision. It aims to facilitate understanding among
practitioners, policy makers, and scholars about
what decentralization entails in practice compared
to theory. A review of the empirical literature and
experience of decentralization is presented. The paper
highlights issues that policy makers in developing and
transitional countries should be aware of when reforming
government, considering their unique political and
economic environment.
The author argues that decentralization produces
efficiency gains stemming from inter-jurisdictional
competition, enhanced checks and balances over the
government through voting at the subnational level,
and informational advantages due to proximity to
citizens. By contrast, arguments against decentralization
include the risk of an increased level of corruption,
coordination problems stemming from multiple layers
of government, low capacity of subnational government,
and unproductive inter-jurisdictional competition.
Decentralization itself does not render increased
government effectiveness in public service provision.
Instead, the effectiveness of government largely depends
on the quality of human capital and institutions.
This paper—a product of the Capacity Building, Partnership, and Outreach Team, Development Economics—is part of a
larger effort in the department to disseminate research findings pertinent to development. Policy Research Working Papers
are also posted on the Web at http://econ.worldbank.org. The author may be contacted [email protected].
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development
issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the
names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those
of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and
its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.
Produced by the Research Support Team
Decentralization and the Provision of Public Services:
Framework and Implementation
Aehyung Kim
Acknowledgements
I am indebted to many individuals. I would like to thank Michael Bell, Martin Lardone,
Garbor Peteri, Rémy Prud'homme, Vito Tanzi, and Serdar Yilmaz for their insightful
comments on earlier drafts, and I am also grateful to Pranab Bardhan, Robert Ebel, Boris
Pleskovic, Steve Webb, Roland White, and Howard Wolman for their helpful discussions
in developing the framework of the paper. All remaining errors are mine and this paper’s
findings, interpretations, and conclusions are entirely those of the author and do not
necessarily represent the views of the World Bank.
2
Introduction
Decentralization has been gaining increasing attention as a means of improving the
provision of public goods and services 1 as reforming public-service delivery has moved
to the forefront of the current policy agenda. 2 The main economic reason behind this
increased interest is the prospect of improving citizens’ welfare through efficiency gains
(Tanzi 1995). In other words, welfare is maximized if each local government provides
the Pareto-efficient 3 level of output for its constituents through benefit pricing (Oates
1972, p. 34-35). The argument is that local government is more efficient to provide
public goods and services tailored toward the preferences and tastes of individual
residents, so as to achieve an increased level of satisfaction for them.
At the same time, experience shows that decentralization has often been
motivated by political reasons (e.g., dealing with a mounting budget deficit) rather than
economic rationale (e.g., efficiency gains, increase in employment, economic
stabilization, equity). For example, an overburdened central government may try to
delegate or pass on the responsibility for the provision and maintenance of public goods
and services to local government, but without ensuring adequate institutional and
financial provision. In this case, decentralization neither improves welfare or the quality
of life for residents nor increases the efficiency of the public sector.
High expectations from decentralization seem to stem partly from recent
experiences of failed centralized states under planned economic and political systems. 4
Strong pressures for the restructuring of the public sector are present today, and have
been before in countries at differing stages of economic development. It has been argued
that fragmenting central authority and introducing more “Tieboutian” intergovernmental
competition (e.g. “voting with feet”) can generate a range of benefits including making
government more efficient and responsive. 5 Some economists and political scientists
also have questioned the effectiveness of central government in improving the
distribution of income and stabilizing the economy, while advocating a smaller public
sector by giving markets and local jurisdictions more power. These arguments have been
put forward along with support for privatization to improve the efficiency of the public
sector, 6 although it is not evident that decentralization can improve income distribution
1
See John G. Head (1974) for discussion on public goods and welfare.
See the World Bank publications for discussion on reforming public service delivery: World
Development Report (2004); Berlin Workshop Series (2004); Annual Bank Conference on Development
Economics-Europe (2005) and Besley and Ghatak (2003).
3
Pareto-efficiency refers to the situation when nothing could be done to make at least one person better off
without hurting someone else. The implication is that no one needs to be made worse off, while at least
one person will have an increased level of welfare.
4
The World Development Report 1997 (World Bank) argues that development requires an effective state
that plays a facilitator role in encouraging and complementing the activities of private businesses and
individuals versus reducing government to a minimalist state. Governments with both centrally-planned
and mixed economies are shrinking their market role because of failed state interventions.
5
Chareles M. Tiebout (1956) argues that the Musgrave-Samuelson analysis, which is valid for federal
expenditures, need not apply to local expenditures.
6
Vito Tanzi, 1996, “Fiscal Federalism and Decentralization: A Review of Some Efficiency and
Macroeconomic Aspects,” in Annual World Bank Conference on Development Economics 1995, pp. 296.
2
3
and macro management. Cai and Treisman argue that “decentralization is sometimes
presented as almost a functional equivalent of privatization.” 7
This paper examines the conceptual framework and experience surrounding
decentralization. It aims to shed light on the effectiveness of decentralization in
improving the provision of public services. First, it begins by presenting definitions of
decentralization. Second, it discusses some of the arguments for decentralization, with
allocative efficiency one of the main economic rationales for improving public welfare.
Third, issues related to the implementation of decentralization are addressed. These
include economic factors influencing the assignment of expenditure and revenue
responsibilities. Some of the major revenue sources are also discussed including own
revenue and intergovernmental transfers. Fourth, the potential impact of decentralization
and its implications for developing countries are reviewed by examining experience with
decentralization. Finally, the conclusion highlights the main points of each of these
sections.
What Is Decentralization?
When discussing decentralization, it is implicitly assumed that countries have different
levels of government (central, regional/state, municipal/local) with different degrees of
decision-making power. Some political scientists argue that each level of government in
a federal system should have independent authority: on this definition, only a few
countries would be considered to operate in a federal system (e.g., US, Canada,
Switzland). Meanwhile, the literature on decentralization does not uniformly agree on
what decentralization entails.
The following are some definitions of decentralization that are broadly coherent,
but have different emphases:
1.
Wallace E. Oates argues that a federal system of government is the most
promising form for organizing the public sector in resolving the allocation, distribution,
and stabilization problems. There are two near polar forms of government:
• Complete centralization (a unitary form of government): in the absence of
other levels of government, the central government assumes full
responsibility for the three economic functions of the public sector.
• Total decentralization: the opposite direction of centralization on the
spectrum of governmental forms that would presumably represent a state of
anarchy. A highly decentralized system in which the central government is
almost completely devoid of economic responsibility. In this case, a system
of small local governments would perform virtually all the economic tasks of
the public sector. But this would be highly unrealistic.
Oates sees that a unitary government has a comparative advantage in resolving
stabilization and distribution problems (including externalities of collective goods) while
7
Cai, Honbin and Daniel Treisman, “Did government decentralization cause China’s economic miracle?”
World Politics, forthcoming, pp.35.
4
a decentralized government has comparative advantages in satisfying varying preferences
of the residents of the different communities by providing public goods and services
whose benefits are limited to a specific subset of the population. Oates argues that
federalism combines the advantages of these two polar forms and avoids the most serious
shortcomings of each. 8 In a federal system, each level of government makes decisions
concerning the provision of certain public services in its respective geographical
jurisdiction: each does what it can do best rather than attempting to perform all the
functions of the public sector.
Oates argues that economists might be interested in which “different levels of
decision-making do exist, each of which determines levels of provision of particular
public services in response largely to the interests of its geographical constituency”
(Oates 1972, pp. xvi). Thus, economists’ central concerns are the allocation of resources
and the distribution of income within an economic system. By this definition, any fiscal
system/government at least possesses federal elements. In that decentralization (fiscal
decentralization) has compelling advantages for some functions over a complete reliance
on centralized decision-making.
In Oates’ view it makes little difference to the economist whether or not decisionmaking at a particular level of government is based on delegated or constitutionally
guaranteed authority in contrast to the conception of federalism in political science. Thus
federalism should be treated not in absolute but in relative terms, in degrees rather than
in kind.
2.
Prud’homme (1994) suggests a typology that distinguishes between three types of
decentralization: spatial, market and administrative. Spatial decentralization is defined
as a process of diffusing urban population and activities away from large agglomerations;
market decentralization, economic liberalization, is defined as a process of creating
conditions in which goods and services are provided by market mechanism; and
administrative deconcentration is defined as the transfer of responsibility for planning,
management, and the raising and allocation of financial resources from the central
government and its agencies to field units of government agencies, or subordinate units or
levels of government.
According to Prud’homme, administrative decentralization can be subdivided into
three types: de-concentration, delegation, and devolution.
• De-concentration is the redistribution of decision-making among different
levels within the central government;
• Delegation is the transfer of responsibilities from the central government to
semi-autonomous organizations not wholly controlled by the central
government but ultimately accountable to it; and
• Devolution is the transfer of powers from the central government to
independent sub-national governments. However, this typology does not
8
Oates interprets Kenneth C. Wheare’s definition of federalism: a political system provided by the
constitution that protects the autonomy of different levels of government, “...the method of dividing powers
so that the general and regional governments are each, within a sphere, co-ordinate and independent (Oates
1972, pp.16).” A political study of federalism is concerned with the division and use of power, by
excluding from the federal category a system in which the power of sub-national governments is exercised
solely at the convenience of the central government: delegation.
5
seem to hold always: the word “decentralization” is often being used for a
synonym for “devolution.” 9
3.
Vito Tanzi ( 1996) describes decentralization in two broad categories:
• De-concentration: implies only a delegation of administrative control to
lower levels (sub-national governments in the administrative hierarchy) and
• Decentralization: genuine possession of independent decision-making power
by decentralized units.
4.
Richard M. Bird and Francois Vaillancourt (2006) consider three types of
decentralization:
• De-concentration: giving regional or local offices of the central government
decision-making power previously held in the central offices in the capital;
• Delegation: making a sub-national government responsible for carrying out a
function for which the central government retains responsibility; and
• Devolution: transferring responsibilities from the central government to subnational governments.
5.
James Manor (1999) considers three types of decentralization: de-concentration or
administrative decentralization, fiscal decentralization, and devolution or democratic
decentralization. 10
• De-concentration/administrative decentralization: refers to the dispersal of
agents of higher levels of government into lower level arena. However when
de-concentration (even with fiscal decentralization) occurs in isolation
without simultaneous democratization, it still enables the central authority to
penetrate more effectively into lower level arena without increasing the
influence of interests at those arena. This could be especially true in
developing countries.
• Fiscal decentralization: refers to downward fiscal transfers to lower levels
influencing over budgets and financial decisions. Fiscal decentralization in
isolation without democratization would face the same problems mentioned
above (e.g., central control) although it is less prone to these tendencies than
de-concentration.
• Devolution/democratization: refers to the transfer of resources and power
(and tasks) to lower level authorities which are largely or wholly independent
of higher levels of government, and which are democratic in some way and to
some degree.
6.
Cai and Treisman illustrate three types of decentralization based on examination
of common usage.
• Administrative decentralization occurs when subnational agents are permitted
by national authorities to make certain policy decisions, subject to review.
9
Rémy Prud’homme (1994) pp. 2.
Manor excludes other three types of decentralization, namely, decentralization by default (when
government institutions become so ineffective, fail to influence lower level arenas), privatization, and
delegation (of responsibilities for development programs or projects to para-statal agencies).
10
6
•
•
Political decentralization occurs when either: (1) subnational governments
have the right to make certain policy decision not subject to being overruled
by higher levels, or (2) subnational officials are chosen by local residents.
Fiscal decentralization implies various forms of fiscal allocation to
subnational governments depending on a specific government.
In summary, three general types of decentralization are presented: administrative,
fiscal, and political.
• Administrative decentralization takes place when funds are allocated to
decentralized units that carry out their spending activities even if most taxes
are raised centrally. 11 It should be noted that in many developing countries,
subnational governments act as agents of the central government according to
guidelines imposed by the central government rather than making
independent spending decisions.
• Fiscal decentralization exists when sub-national governments have decisionmaking power, authorized by the constitution or legislation, to raise revenues
(taxes) and perform spending activities. 12
• Political decentralization exists when public officials (e.g., governor, mayor,
council member) at sub-national governments are elected by secret ballots,
and sub-national governments are given independent power for decisionmaking by constitutional or legislative authority.
Table 1: Types of decentralization and their characteristics
Type
Administrative
Independent
Characteristics decision-making by
subnational
governments on
spending activities
Fiscal
Independent decisionmaking power by subnational government
for raising revenues
and allocation of
financial resources
Political
Free election in
selecting officials (e.g.,
mayor, council) of
subnational
governments by secret
ballots
In practice, decentralization is often used with interchangeably devolution, and
de-centralization with delegation. It also should be noted that decentralization, in a
theoretical sense (and in an ideal world), requires the simultaneous occurrence of all these
three types of decentralization (administrative, fiscal, and political) in order to realize the
full potential benefits. However, in the real world (and more so in developing countries)
each of these can occur in isolation, or any two (or all) can occur simultaneously to
different degrees. 13 Thus, decentralization can be better described as an evolutionary
11
Italy could be considered as an administratively decentralized country where local entities raise a small
amount of total revenue but spend a relatively large portion of total net expenditure (Vito Tanzi 1996, pp.
297).
12
Examples of fiscal decentralization include Argentina, Australia, Brazil, Canada, India, Germany,
Nigeria, Switzerland, and the United States (Vito Tanzi 1996, pp. 297).
13
James Manor (1999).
7
process for allocating administrative, fiscal, or political power to subnational
governments, rather than being defined precisely by these specific features.
For example, in measuring fiscal decentralization, the most widely used indicator
is the population weighted average share of subnational expenditure and revenue in the
total public sector, as reported in the Government Finance Statistics (GFS) of the
International Monetary Fund. The average expenditure and revenue shares of subnational
governments in twenty-eight countries have been increasing steadily over time since
1980. However, GFS is not an ideal data set for measuring fiscal decentralization:
although it provides a breakdown of expenditures by function and economic type, it lacks
information on the degree of expenditure autonomy. On the revenue side, GFS does not
provide information on whether taxes are collected through shared taxes, piggy-backed
taxes, or locally determined “own-source” taxes, or what proportion of intergovernmental
transfers is conditional or unconditional (Ebel and Yilmaz 2002).
Another effort is underway by the Organization for Economic Cooperation and
Development (OECD) to measure the degree of subnational autonomy over expenditure
decisions as well as own-revenues and grants (see Bell and Ebel 2006). Roy Bahl sums
up that “fiscal decentralization refers to an intergovernmental system where the balance
of power moves more toward the subnational government sector than has been the case.”
14
Why Decentralization?
In recent years, many countries around the world have begun to undertake
decentralization in response to “government failure” in delivering public goods and
services. It is viewed as a way to make government more efficient, responsive, and
accountable. These perceived benefits of decentralization have attracted a diverse range
of supporters 15 that favor small government, free-markets, and a bottom-up approach,
while privatization is encouraged as a way to advance decentralization (Tanzi 1996,
Bardhan 2002). Before discussing the rationale for decentralization, a brief note on
efficiency regarding the provision of public services is in order since the government’s
resource allocation policies (e.g., revenue and expenditure) influence it. First, two
dimensions of efficiency are discussed: productive and allocative. Second, arguments for
improving welfare through government allocation are considered. 16
Productive efficiency is when the economy is working on its production
possibility frontier: this is when production of one good is achieved at the lowest cost
possible, given the production of the other good(s). 17 In other words, it is the highest
possible output of one good, given the production level of the other good(s). Meanwhile,
allocative efficiency is the market (in this case the public sector) condition whereby
14
Bahl (1999, pp. 4)
Pranab Bardhan calls them “anarcho-communitarians” (2002, pp. 186).
16
Allocation function is one of the three major economic functions of government identified by among
others, Musgrave (1956), Oates (1972); the other two functions are the distribution function and the
stabilization function.
17
A production possibilities curve (PPC) or “transformation curve” is a graph that shows the different
quantities of two goods that an economy can produce if it uses its limited productive resources efficiently.
Points along the curve describe the trade-off between the two goods. This curve shows the possibilities
open for increasing the output of one good by reducing the output of another.
15
8
resources are allocated in a way that maximizes the net benefit attained through their use.
Allocative efficiency refers to a situation in which the limited resources of a country are
allocated in accordance with the wishes of consumers. An allocatively efficient economy
produces an "optimal mix" of commodities. Resources are used allocatively efficiently if
they are spent for producing the right mix of goods for the right people at the right price.
That is to say, in a situation where society produces goods and services at minimum cost
that satisfy the preferences of consumers, resources are allocated efficiently.
A set of rules devised 18 in welfare economics if fulfilled will give rise to a Pareto
Optimum, i.e., no one can be made better off without someone else being worse off. In
other words, if somebody could be made better off without making any other person
worse off, then clearly welfare is not maximized, and therefore resources are not
allocatively efficient. If an economy is perfectly competitive, it is assumed that
equilibrium exists, i.e. Pareto-efficiency, based on individual welfare.
However, Pareto-efficiency does not ensure socially desirable objective such as
equity. In addition, Pareto-efficiency based on individual welfare may not also be
efficient because of externalities since individuals take account only direct effects on
themselves, not the effect on others. Furthermore, a competitive economy may not
ensure a Pareto-efficient allocation of resources due to “imperfect information.” The role
of government is thus to ensure the correct supply of public goods while its intervention
in markets is justified on the ground of correcting “market failures.” 19
Furthermore, unlike the private sector which has the single goal of maximizing
profits for its share holders, the public sector’s goal is to increase the benefit of the public
at large—even though preferences are diffuse and welfare may be hard to measure. In
this sense, Bergson (1983) suggests that a “Pareto-improvement” should be concerned
with “normative efficiency” in formulating optimization criteria. 20
In the context of decentralization, a Pareto efficient allocation of resources alone
may not necessarily lead to the optimal allocation of resources to meet the collective
interest of citizens in a particular jurisdiction. Because Pareto efficiency does not
necessarily produce public goods and services that reflect the desirable characteristics of
these services, simply that no one person could be made better off without making some
other person worse off. There is no consideration, for example, for equity. Thus,
allocative efficiency occurs when the public sector produces public goods and services
most valued by society. This means scarce resources are allocated to the production of
goods and services so that citizens’ wants and needs are met in the best way possible. 21
As alluded before, the role of the public sector is to achieve a welfare optimum by
allocating resources efficiently and facilitating economies i.e., through providing
necessary public goods and services in promoting job growth, fair income distribution,
18
The basic theorems of welfare economics require that the economy is perfectly competitive and there is
a full set of markets (e.g. identical marginal rate of substitution in consumption, identical marginal rate of
transformation in production) and a number of market conditions (e.g. asymmetric information,
externalities, natural monopoly, and principal and agent problems) may not lead to inefficiency.
19
See Joseph Stiglitz (1980) for detailed discussion.
20
For further discussion see: Abram Bergson (1954, 1983) for social welfare; John Rawls (1971) for “maxmin” utility function concerning equity and social economic justice; Amartya Sen (1984) for equity and
poverty.
21
See Joseph Atkinson, A. B. and Joseph Stiglitz (1980) for detailed discussion on issues of efficiency and
the economics of the public sector.
9
and efficient allocation of resources (Musgrave 1959, Oates 1972). Although these three
functions are conceptually distinctive, they are fundamentally interdependent and
moreover, at a policy level, any public problem will have an impact on the attainment of
all three of the objectives of public policy.22
However, there exist different levels of decision-making in a federal structure 23 of
the public sector; each of which determines the provision of particular public services in
responding largely to the interests of its geographical constituency (Oates 1972). An
ideal case, “the optimal form of federal government to provide the set of n public goods
would be one in which there exists a level of government for each subset of the
population over which the consumption of a public good is defined.” 24 Oates calls this
ideal case “perfect correspondence” in the provision of public goods in which a perfect
match exists between the jurisdiction of each government unit and the group that
collectively consume the goods. 25 For example, special districts in the U.S. reflect this
concern such as fire, sewerage, and water districts. However, it should be noted that it is
implicitly assumed that the costs of providing each level of output of the good 26 in each
jurisdiction are the same for the central or the respective local government.
For example, central government tends to be better equipped at dealing with
stabilization and distribution problems (e.g., monetary and fiscal policy) in maintaining
the economy at the national level, while local governments are constrained in their
capacity to regulate the aggregate level of economic activity. On the other hand, local
governments may be more efficient, for example, in providing public goods of a local
nature (e.g., refuse collection) for their constituency. They have comparative advantages
over central government to provide the mix of public goods and services that best reflect
the preferences of individuals in the jurisdiction since they have better access to
information on local affairs. While a centralized government might not be well informed
about them or ignore the difference, and would thus tend to supply a uniform package to
all residents, and does not deliver the Pareto-optimal level of satisfaction for them. Thus
the main economic argument for decentralization is largely based on allocative efficiency
grounds (Tanzi 1994).
According to Oates, “it will always be more efficient (or at least as efficient) for
local government to provide the Pareto-efficient levels of output for their respective
jurisdictions than for the central government to provide any specific and uniform level of
output across all jurisdictions.” 27 The argument is that welfare is maximized if each
local government provides the Pareto-efficient output for its constituency based on the
principle of “perfect correspondence.” Olson (1969) also argues that there is allocative
efficiency where the boundaries of a government and a collective good it provides
22
Wallace E. Oates, 1972, Fiscal Federalism,
Oates (1972) argues that “any fiscal system is federal or at least possesses federal elements” and “it is not
surprising, for decentralized finance has such compelling advantages for some functions that one would not
expect to find a total reliance on centralized decision-making on fiscal matter (pp. xvi).”
24
Oates (1972) pp. 34.
25
Oates (1972) pp. 34.
26
If this is not the case, for example, if the central government can realize important economies of scale
(scope) in the provision of the good that are not available to the local government individually,
centralization may be desirable despite the decentralized character of the consumption of the good (Oates
1972), pp. 37.
27
Oates (1972) pp. 35.
23
10
coincide. “There is a need for a separate governmental institution for every collective
good with a unique boundary, so that there can be a match between those who receive the
benefits of a collective good and those who pay for it (pp. 483).”
In addition, the incentives for decentralization are likely to be greater the more
varied the efficient levels of output across jurisdictions; for the more pronounced will be
the divergence of any uniform level of output from the efficient level in most localities.
For a given population size, the welfare gain from the decentralized provision of a
particular local public good becomes greater as the diversity in individual demands
within the country as a whole increases and as each geographical grouping of consumers
becomes more homogeneous in terms of their demands for the good (see appendix for
graphical presentation).
At the same time, the “subsidiarity principle” 28 argues that the efficient provision
of public goods and services requires that the government satisfy citizens’ needs and
preferences as much as possible. Taxing, spending, and regulatory functions should be
exercised by lower levels of government unless a convincing case can be made for
assigning them to higher levels of government. This principle is the opposite of the
residuality principle applied in a unitary country, where local governments are assigned
functions by the central government. Thus the subsidiarity principle and decentralization
in federalism should produce the optimal allocation of resources and consequently
maximize welfare.
Implementing Decentralization
As shown from experience in decentralization, it is rarely the case all of the three types of
decentralization occur simultaneously. Instead, it is more common to see one or two
aspects of decentralization implemented in many countries; when governments
decentralize they are often really deconcentrating or delegating rather than devolving
responsibility (Martinez-Vazquez, McLure, and Vaillancourt 2006). At various degrees
and scopes of decentralization, countries around the world have attempted to expand the
role of sub-national governments in providing public goods and services for their
citizens. 29 Most subnational governments deal with fiscal activities for generating
revenue and making decisions on spending (expenditure) for the provision of public
goods and services.
a. Assigning Functional and Expenditure Responsibilities
As mentioned in the previous section, the division of expenditure and revenue
responsibilities and revenue sources is largely determined by how government activities
are assigned to different levels of government. At the same time, assignments of
functional responsibilities depend on the relative competence of different levels of
government for carrying out a particular functional area. The three major functions of
government include: macroeconomic stabilization, income redistribution, and resource
28
Martinez-Vazquez et al (2006, pp. 18) states that “this is put forward as a key principle in the proposed
European constitution (Article 1-9).”
29
Roy Bahl (1999) does not see “true fiscal decentralization” would be viable in developing countries at
least in the short run, given their very limited tax base at the local level.
11
allocation. Four criteria can be used to assess the assignment of functional and
expenditure responsibilities.
Four Criteria. At the same time, each level of government is responsible for
activities based on four (not exhaustive) criteria: economies of scale; the presence of
externalities; heterogeneity of preferences and of circumstances; and emulation. 30
• Economies of scale: vary across goods and services and size of jurisdiction.
For example, the unit cost per good drops as the quantity of the good
increases. The existence of significant economies of scale constitutes an
argument for a higher level of government to provide a particular good or
service.
• The presence of externalities (negative or positive): provision of a good or
service may have spill-over effects beyond a particular jurisdiction. The
presence of external effects is a consideration in determining the optimalsized group to serve a particular good. For example, the smaller the group,
the less the external effects are likely to be internalized. The greater the
externalities (as aggregate welfare gain or loss becomes larger), a higher level
of government provision is desirable. 31
• Heterogeneity of preferences and of circumstances: when there is the
existence of heterogeneous preferences among geographical subsets of the
total population, welfare would be maximized if each local government
satisfies the unique preferences of its citizens by providing a particular good
or service tailored to citizens’ preference (e.g., quality) instead of uniform
provision of a good across the region or country by the central government.
• Emulation: refers to competition, which facilitates introducing best practices
in government, requires two or more jurisdictions involved in a given
activity; an argument for decentralizing government activities.
b. Two Major Sources of Subnational Government Revenue: Local Taxes and
Intergovernmental Transfers
Governments carry out their responsibilities through spending. Expenditures
should be matched by revenues. Revenues basically consist of three sources: local tax
revenues, intergovernmental transfer, and to some extent, borrowing. Borrowing is
certainly another option, but recent studies (Tanzi 1996; Rodden, Eskeland, and Litvack
2005) suggest this may be problematic because of the lack of hard budget constraints
imposed on subnational government in most developing countries (and in many OECD
countries). Central government bail-outs are rather common and the weak enforceability
of hard budget constraints of subnational governments puts macro-economic stability at
risk (Tanzi 1996, Bardhan 2002, Tommasi 2006). This is particularly a problem for
developing countries, except perhaps for a few wealthy metropolitan cities.
30
Martinez-Vazquez, McLure, and Vaillancourt (2006).
Oates (1972, pp. 46) adds “free-rider” problems when the provision of the good elsewhere can serve a
substitute for local provision of a good. Then there will be an incentive to engage in “free-rider” behavior
(accentuates with a large number of small jurisdictions of consumers).
31
12
However, only two sources will be discussed in this paper. Local tax revenues
typically include income taxes (piggy backed to federal income taxes in many developing
countries due to collection difficulties), property taxes, user fees, license, excise fees.
Intergovernmental transfer (reflecting disparities between central and subnational
governments and across different subnational governments) can be conditional (e.g.
matching grants) or unconditional (revenue sharing) grants from the central government.
Depending on the degree and scope of decentralization, sub-national governments either
have autonomy deciding on what to and how to spend money on goods and services or
they have to follow guidelines imposed by the central government. However, it should
be noted that in most countries subnational governments have very limited tax bases and
capacity to set tax rates and administer the collection of taxes. As a result, the central
government determines the tax base and rates as well as collecting centrally. Thus,
intergovernmental transfer is the most important resource for subnational government.
b-1. Own Revenue Sources of Subnational Government
The Principle of Benefit Taxation. Subnational governments raise their own
resources from various taxes and fees. If subnational governments are to achieve true
fiscal decentralization, they must control their own sources of revenue. Thus important
questions are which revenue sources can and should be assigned to subnational levels of
government and how these assignments will be carried out. The principle of benefit
taxation 32 represents that to the extent possible, goods and services provided by the
government should be financed by fees and taxes. Olson refers to “fiscal equivalence” in
which there is a match between those who receive the benefits of a collective good and
those who pay for it (1969). Tax payments should reflect the costs and benefits of public
goods and services. This is fair and efficient in the sense of encouraging responsible use
of the nation’s scarce resources. Such taxes levied on motor vehicles and fuels are some
of the examples of benefit-related taxes used for the construction and maintenance of
roads.
The assignment of taxes should match the benefits derived from spending based
on assigned responsibilities and expenditures of each level of government. Approaches
to assigning revenues to subnational governments differ in the degree of fiscal autonomy
in terms of the ease of compliance and administration, the fairness and neutrality they
can likely produce, and the degree of interjuridictional redistribution they can
accommodate. According to Martinez-Vazquez et al (2006), four aspects of revenue
assignment can be distinguished: (1) which level of government chooses the taxes from
which subnational governments receive revenue; (2) which level defines the tax base; (3)
which level sets the tax rates; and (4) which level administers the taxes. For the view of
subnational fiscal sovereignty, the authority to set rates is the most important of these
because the choice of rates is what allows subnational government to choose the level of
public services. However, excessive subnational latitude in the choice of tax bases and in
tax administration can cause unacceptable administrative burdens and distortions in the
allocation of resources.
32
The principle of benefit taxation refers that within the context of the just distribution of income,
consumers finance the cost of public goods and services by payments in accordance with benefits received.
13
Approaches for Subnational taxes. In general three approaches are practiced.
Under a first approach, subnational governments have fiscal autonomy provided by
independent subnational legislation and administration in choosing taxes they levy, tax
bases, setting the tax rates, and administering the taxes. This approach is practiced in
Canada, Switzerland, and the United States. Under a second approach, a higher level of
government defines the tax base and collects both its own and surcharges set by
subnational governments. With this approach, subnational governments retain the
attributes of fiscal sovereignty in the tax field by setting the tax rate rather than retaining
the ability to define the tax bases and administer taxes. Under a third approach (tax
sharing), subnational governments receive fixed fractions of revenues from particular
national taxes originating within their boundaries. However, this approach restricts
subnational governments’ fiscal autonomy even if they have authority over spending a
given amount of revenue (see Tommasi 2006 for extensive discussion on tax sharing in
Argentina during the 1990s).
Based on the traditional theory of fiscal federalism (easy to administer locally and
mainly imposed on local residents), subnational governments are usually given a limited
tax base such as property tax, taxes on vehicles, and user charges and fees. These limited
sources of revenue impose difficulties on subnational governments in financing necessary
public services. For example, property tax is not sufficient to meet major social
expenditures in most cases; in addition, it is costly and difficult to administer. 33 Richard
Bird (2006) argues that good subnational taxes should satisfy two main criteria: (1) they
should provide sufficient revenue for the autonomous fiscal units, at least for the richest
subnational units; and (2) good subnational taxes should impose fiscal responsibility at
the margin on subnational governments by allowing them to establish their own tax rates
with respect to at least some major taxes.
Major Tax Sources for Subnational Government. The following are the major
tax sources for subnational governments (local and regional).
• User charges: these are based on the principle of “willingness to pay” rather
than the principle of “ability to pay.” Citizens should value what the public
sector supplies, at least at its marginal cost. It is a way to assess the true
demand for the goods or services the public sector supplies, thus the
economic efficiency in allocating resources can be enhanced. However, the
importance of user charges is greater in principle than the relatively small
amount of revenues collected from this source in most countries.
• Property taxes: in spite of difficulties in assessing the value of property and
administrative costs, a low-rate and uniform property tax is important in
financing local governments in most countries. Subnational governments
usually finance “hard” services such as local roads and garbage collection out
of property taxes. However if these governments are to expand their roles
into “soft” services such as education and health, then they will need to have
access to more elastic revenue sources.
• Excise taxes: while the property tax is the pre-eminently local tax, excise
taxes (on alcohol and tobacco) could be a significant source of subnational
33
Bell and Bowman (2005) explore a way to extend local property taxation into communal lands in rural
South Africa.
14
•
revenue. Excise taxes are relatively easy to administer and lend themselves to
regionally-differentiated rate determination. In addition, a general benefit
argument can be made for some regional excises, for example, on alcohol and
tobacco to the extent that subnational governments are responsible for health
expenditures and on vehicles and fuel to the extent that they are responsible
for regional and local roads.
Other potential sources of taxes include personal income taxes, payroll taxes,
sales tax, and business taxes. Except user charges to some extent, most of
these taxes (including property and excise) are not likely reduce local and
regional disparities between jurisdictions.
To implement fiscal decentralization, three principles are suggested for
subnational taxation (Bird 2006): (1) attention should be paid to matching expenditure
and revenue needs; (2) an effort should be made to ensure that all governments bear
significant responsibility at the margin for financing the expenditures for which they are
politically responsible; and (3) subnational taxes should not unduly distort the allocation
of resources. In many developing and transition countries, however, problems related to
subnational taxation would largely remain unsolved in the foreseeable future even if there
are major reforms of subnational tax regimes, due to limited tax bases and administrative
capacity at the subnational level. That is why many transition countries have opted for
shared taxes.
b-2. Intergovernmental Transfers
Regarding intergovernmental transfer, several experts (Oates 1972, Bhal 1999)
argue that unconditional grants are preferable to conditional grants in promoting
decentralization. In the case of conditional grants, the central government defines to
some extent, the objectives for which sub-national governments are to use the grants in
promoting positive externalities across regions (e.g. education, health programs, interregional transportation network). On the other hand, with unconditional grants,
subnational governments can employ the grants according to their own set of priorities
and objectives. But, there is an indication that conditional grants (e.g. matching grants)
that earmark to infrastructure are more likely to promote infrastructure investment (Oates
1972) more than non-conditional grants. It might also be desirable if the central
government wants to promote positive externalities generated by multi-jurisdictional
infrastructure projects through this kind of subsides (“Pigovian” subsidies). Then, the
central government may want to use conditional grants as a policy tool in creating certain
incentives for sub-national governments.
Unconditional grants, to some extent, imbed an element of the welfare optimum
involving both allocative efficiency and an equitable distribution of income (Oates 1972,
pp. 78). In Australia, Canada, and several European states, the central government
makes substantial unconditional grants to subnational governments containing significant
equalizing features. 34 At the same time, a national program that distributes income
34
Based on equal per-capita grants to all states, the grants are usually financed by the progressive federal
income tax which wealthier jurisdictions would pay in a larger sum per capita to support the grant program
although they receive the same per-capita grant as poor states.
15
among individuals, not among jurisdictions, is the preferred alternative to achieve a just
distribution of income among individuals in a country. Unconditional grants are also
used for the promotion of horizontal equity (the equal treatment of equals) 35 among
jurisdictions. The central government seeks to equalize the fiscal capacity of all
jurisdictions by instituting a transfer program of unconditional grants to poorer localities.
For example, grant per capita can vary inversely with the level of per-capita income in
the jurisdiction.
Most countries use both conditional and unconditional grants to more
decentralized levels of government. The former are typically used to encourage spending
on such items as education, health, highways, and regional roads, which involve
significant interjurisdictional positive externalities, in addition to support explicitly
redistributive programs. The major conditional grant programs intend to achieve two
primary objectives: (1) the achievement of national minimum levels of certain basic
public services; and (2) an equalization across jurisdictions of the “fiscal effort” required
to provide these minimum program levels. It is to ensure that every individual has access
to a sound program of public services within his chosen locality at a cost in line with
what he would pay elsewhere.
Moreover, these intergovernmental grants in general involve equalizing
provisions (e.g., horizontal equalization) so that poorer jurisdictions receive more
generous support by incorporating provision to account for the particular expenditure
requirement, or “need” of each jurisdiction in addition to its “fiscal capacity” to meet that
need. Intergovernmental transfer thus can be interpreted as a policy tool by the central
government in achieving one of its prime responsibilities in achieving income
redistribution. In addition a good intergovernmental transfer system should provide fiscal
incentives for local governments to economize on expenditures and to raise their own
resources.
Challenges and Potential Impacts and of Decentralization
with Implications for Developing Countries
Policymakers, researchers and the public have long been interested in ways to increase
the effectiveness of the public sector. The core of the debate lies in how government can
improve the welfare of its citizens through adequate provision of public services, from
opportunities for education and employment to the provision of clean water.
Decentralization has been seen by a wide range of advocates as a way to increase the
efficiency and enhance the accountability of the public sector, i.e. improving the
governance of the public sector. These potential benefits have engendered an enthusiasm
for decentralization in both developing and developed countries. Developing countries
are turning to decentralization to escape from the traps of ineffective public sectors, while
developed countries look for effective tools to reorganize the government in order to
efficiently provide public services in the “post-welfare state” era (Bennett 1990; World
Bank 1997; Bird and Vaillancourt 2006).
35
Musgrave (1959) argues in The Theory of Public Finance (pp. 160) that “perhaps the most widely
accepted principle of equity in taxation is that people in equal positions should be treated equally.” Grants
are also provided to alleviate vertical fiscal disequilibrium between different levels of government by
assessing the needs based on such indicators as population, urban density.
16
However, decentralization is neither good nor bad. “It is a means to an end, often
imposed by political reality (World Bank 1999, pp. 107).” In addition, it is increasingly
clear that the benefits of decentralization can easily be undermined by potential costs
such as increasing indebtedness, high deficits, and procyclical finances of subnational
governments and regional disparities (Wolman 1994; Prud’homme 1994, 1996; Tanzi
1996; Rodden, Eskeland, and Litvack 2005; Bardhan and Mookherjee 2005; Tommasi
2006). At the same time, the potential allocative efficiency gains of decentralization
stemming from satisfying unique preferences of citizens in jurisdictions may not be
realized as assumed in theory. Because in most developing countries, the problem is not
in satisfying fine differences in preferences between jurisdictions but to meet basic needs.
a. Accountability
Regarding public sector accountability, it is argued that the accountability of the
public sector will be improved because ineffective local officials would be voted out
from their office because the taxpayers will express their preferences in their votes.
However experience shows that local elections are often dictated by personal loyalties or
political party loyalties. Even if taxpayers vote based on their preferences, it is often the
case that voters have limited access to relevant information (“asymmetric information”)
and often limited capacity to digest the information (see Sappington and Stiglitz 1987 for
asymmetric information; Downs 1957 for voting in democracies 36 ). Devarajan et al
(2007) show that the “missing link” between citizens and local government officials in
the provision of public services is due to the lack of enforceability of local officials’
accountability. This problem tends to be more severe in developing countries.
At the same time, there is no guarantee that locally elected officials will fulfill
their obligation to satisfy the revealed preference. Incentive systems are such that the
officials may satisfy a number of influential people in business and political communities
(“local elites”) instead of their constituency in order to be re-elected. In particular if reelection depends largely on the influence of local leaders and interest groups rather than
on the local performance of public services, as it usually takes longer than the election
cycle to produce tangible benefits with policy changes. Thus “local capture” can easily
undermine the benefits of decentralization by incurring new costs (e.g. poor subnational
tax collection, distortionary subnational taxation, inefficient provision of local public
goods).
Experience on decentralization also points to the effects of corruption at the local
level although there is no conclusive evidence whether centralized or decentralized
government creates higher costs related to corruption (see Bardhan, Pranab and Dilip
Mookherjee 2005 for a detailed survey on corruption). Several studies warn that
corruption is costly in terms of allocative efficiency and production efficiency. For
example, bribes and kickbacks from suppliers would lead to an oversupply of services.
36
Downs (1957) argues that the empirical observation of the economic theory of a democracy suggest that
democracy is an imperfect political system as the government is chiefly motivated to maximize the most
votes instead of maximizing the social welfare of the society. This imposes continuous challenges to
governments and citizens alike as to what degree we can collectively ensure a relatively equal society
where the will of people and the public interest are represented relative to the proportion of the population
and not relative to the endowment of resources, to help ensure economic and political stability.
17
At the same time, productive efficiency would be undermined when cost-saving
technologies and policies are not adopted.
b. Regional Disparities and Economic Development
Decentralization can also increase regional disparities between jurisdictions.
Prud’homme (1994, 1996) argues that decentralized redistribution is likely to lead to
different treatment of similar individuals in a country. For example, jurisdictions with
higher per capita income would be able to provide higher levels of public services than
those with lower per capita income. Residents in wealthier jurisdictions could even be
levied at lower tax rates for higher levels of public services than those in poor
jurisdictions.
Moreover, political fragmentation could hamper the promotion of regional
economic performance and worsen the disparities within it. For example, local
metropolitan governments compete with each other as locations for employers so as to
improve their tax base. This need not increase the aggregate level of employment in
metropolitan areas, it merely redistributes employment. Local governments have
incentives to exclude high cost residents and are reluctant to invest in improving the skills
of the poor. Fragmentation tends to encourage a “beggar thy neighbor” policy (Wolman
H. et al 1992). Therefore equity between localities and individuals should be a concern
when considering decentralization. In many developing countries where income
distributions between jurisdictions and individuals are skewed, equity and fairness should
be an important consideration in designing decentralization.
Experience on decentralized government in the U.S. also raises an issue related to
public policy in promoting development through state enterprise zone programs. Peters
and Fisher (2002) argue that excessive competition between states and localities to attract
business and firms through enterprise zone programs has not produced any noticeable
positive impacts on either job growth or tax revenue. Front-loading incentives for
attracting business through tax rebates and infrastructure provision have failed to provide
any sustained investment growth; instead the venture has wasted tax revenues of the
subnational government. Although subnational governments should take an active role in
promoting economic growth within their jurisdictions, the fragmented government
structure does not seem well equipped to implement coherent economic development
strategies for long term economic growth.
C. Local Capacity
Although their decentralization experience has not resulted in the macroeconomic
instability associated with Latin America, White and Smoke (2005) observe that East
Asian countries face the major challenges in decentralization. One of them is the poor
accountability and weak capacity (both financial and human resources) of subnational
governments. The other two relate to the sound assignment of functions among different
levels of government and concerns for the sufficient allocation of own-source revenues to
subnational governments to ensure the adequate provision of services.
At the same time, Yilmaz, Hegedus, and Bell (2003) emphasize that in order to
realize the full potential of decentralization, the availability of necessary information at
18
the local level is critical. In many developing countries, as local governments take a
more active role in the provision of education and health services (which help reduce
poverty), weak statistical capacity at the subnational level to produce the household data
needed to identify potential beneficiaries and monitoring progress in poverty reduction
prevents full realization of decentralization.
Bardhan (2002) also highlights the need to strengthen the local capacity of
information and accounting systems and mechanisms of monitoring the performance of
public bureaucrats in developing countries. For instance, attention should be paid to
special incentives and devices to check bureaucratic corruption, thus the differential
efficacy of such mechanisms under centralization and decentralization can be evaluated.
In many developing countries, one of the major goals is to effectively reach out to the
poor—targeting success in poverty alleviation programs is treated as an important
criterion, rather than the efficiency of interregional resource allocation. Thus,
redistribution to disadvantaged groups or regions at the subnational level can improve the
potential for productive investment and human resource development on the part of
communities long bypassed by the local elite.
Conclusion
Musgrave (1956) argues that a federal structure for the public sector has the greatest
potential for resolving the stabilization, distribution, and allocation problems while Oates
emphasizes that all forms of government in one way or another share the characteristics
of federalism, regardless of degree of centralization or decentralization. Theories of
federalism and empirical studies suggest that there tends to be an inverse relationship
between centralization and income level, population size, and diversity (cultural, ethnic,
regional, racial). At the same time, many developing countries are searching for an
alternative to the ineffective provision of public services by the central state. Thus, it is
likely that many countries are likely to face issues of decentralization as countries grow
in terms of income and population, and to improve the efficiency of provision of public
goods and services.
First, this paper discussed several definitions of decentralization. It showed that
devolution is often used interchangeably with decentralization, while delegation is often
substituted for de-concentration. In general, decentralization refers to the independent
decision-making authority of subnational government in its administrative, fiscal, and
political affairs. But in reality, most countries are likely to simultaneously deal with
issues related to these two dimensions (e.g., delegation and devolution) of
decentralization, not just one or the other.
Second, the paper presented efficiency arguments for decentralization, by
examining theories of federalism where each level of government has a comparative
advantage in the provision of particular public services. For example, the central
government might be more efficient in dealing with macrostabilization and redistribution
policies. However, subnational government is able to more efficiently provide public
services of a local nature within its jurisdiction, because it has better access to relevant
information and is more accountable for satisfying the preferences of its residents. Thus,
decentralization of certain public services will improve the welfare of a country’s citizens
at the aggregate level.
19
Third, in discussing implementing decentralization, it was argued that the proper
assignment of functional responsibilities depends on the relative competence of different
levels of government for carrying out a particular functional area (e.g., macroeconomic
stabilization, income redistribution, and resource allocation). Four factors (externalities,
economies of scale and scope, heterogeneity of preference, and emulation) are considered
for assigning functional responsibilities. It is argued that certain types of economic
parameters influence the structure and operation of the public sector, although the
structure of a particular country is largely the result of the unique political and social
history of that nation. At the same time, the assignment of government activities
determines the division of expenditure and revenue responsibilities.
Expenditures are financed through subnational own sources of revenue and
intergovernmental transfers. Subnational government’s own sources of revenue include
user charges, fees, and excise and property taxes. These taxes are based on “the principle
of benefit” that users should pay for services provided. Other potential sources of
revenue are retail sales taxes, income taxes, and value added taxes. But these taxes are
usually collected centrally and shared between different levels of government in most
developing countries. Intergovernmental transfers are the other major source of revenue
for subnational government in most countries, due to limited tax bases and weak
administrative capacity in setting tax rates and collecting and administering taxes. In
addition, borrowing is another option, but recent studies (Tanzi 1996; Rodden, Eskeland,
and Litvack 2005; Bahl and Martinez-Vazquez 2006, Tommasi 2006) suggest this may
be problematic because of the lack of hard budget constraints imposed on subnational
government in most developing countries (and in many OECD countries). Broadway and
Shah (2007) argue that subnational governments are almost never self-sufficient
financially—regardless of the political or constitutional definition of the nation—and
have to rely on financial transfers from the central government. Thus, at least in the
short term, a key issue is how to design an effective intergovernmental transfer that
ensures the adequate provision of public services.
Finally, the paper examined challenges and potential impacts of decentralization,
along with their implications for developing countries. In recent years, many developing
countries have taken a keen interest in decentralization to address the problem of an
inefficient public sector. But, experience shows that the negative impacts of
decentralization can undermine the benefits. Some of the challenges for decentralization
include the limited capacity of local government, corruption of local officials, and
imperfect democratic systems. These imply that decentralization may not fully deliver
the intended objectives of improving public sector management and efficiency of
resource allocation in the provision of public services.
However, there is growing pressure to find an alternative to the ineffective
centrally controlled provision of public services in many developing countries even if
decentralization is not a panacea for correcting inadequate provision of public services.
In order to fully take advantage of the potential benefits of dececentralization, Bahl
(1999) suggests twelve rules that need to be in place for the viable implementation of
fiscal decentralization. 37 For the first rule, Bahl argues that implementation should begin
with a design of a comprehensive system and be prepared to deal with the transition
problems during phase-in. In addition, he considers that political autonomy is the most
37
See Bahl (1999) for detailed discussion on twelve rules for implementing fiscal decentralization.
20
crucial element of a decentralized system, along with a significant set of expenditure
responsibilities and a substantial amount of taxing powers, budget making autonomy,
transparency, and a hard budget constraint. At the same time, it is crucial for many
developing countries to improve institutional design in order to reshape existing forms of
organization of the public sector and improve the provision of public goods and services.
Therefore, issues of accountability and incentives are as important as issues of optimal
taxation, tax incidence, and so on.
21
References
Atkinson, A. B. and Joseph Stiglitz. 1980. Lectures on Public Economics, McGraw-Hill
Education.
Bahl, Roy. 1999. Implementation Rules for Fiscal Decentralization, Georgia State
University, Atlanta, Georgia.
Bahl, Roy and Jorge Martinez-Vazquez. 2006. Sequencing Fiscal Decentralization,
Policy Research Working Paper, no. WPS 3914, the World Bank.
Bardhan, Pranab. 2002. “Decentralization of Governance and Development,” Journal of
Economic Perspectives, Fall, Vol. 16, No. 4, pp. 185-205.
Bardhan, Pranab and Dilip Mookherjee. 2005. “Decentralization, Corruption and
Government Accountability: An Overview,” in Handbook of Economic
Corruption, edited by Susan Rose-Ackerman and Edward Elgar.
Bartik, J. Timothy. 1991. Who Benefits from State and Local Economic Development
Policy? pp. 202-211.
Bell, Michael E. and John H. Bowman. 2005. Implementing A Local Property Tax When
There Is No Real Estate Market: The Case of Commonly Owned Land In Rural
South Africa.
Bell, Michael E. and Robert D. Ebel. 2006. Statistical Capacity Building in Tanzania and
Uganda: Identifying and Measuring Own-Source Subnational Revenue Autonomy.
Bennett, Robert J. 1990. “Decentralization, Intergovernmental Relations and Markets:
Towards a Post-Welfare Agenda?” in Bennett (editor) Decentralization,
Intergovernmental Relations and Markets: Towards a Post-Welfare Agenda,
Clarendon Press, Oxford.
Bergson, Abram. 1954. “On the Concept of Social Welfare,” The Quarterly Journal of
Economics, Vol. 68, No. 2, pp. 233-252.
--------------------. 1983. “Pareto on Social Welfare,” Journal of Economic Literature, Vol.
21, No. 1, pp. 40-46.
Besley, Timothy and Maitreesh Ghatak. 2003. “Incentives, Choice, Accountability in the
Provision of Public Services,” Oxford Review of Economic Policy, Vol. 19, No. 2
Bird, Richard M. 2006. “Fiscal Flows, Fiscal Balance, and Fiscal Sustainability,” in
Perspectives on Fiscal Federalism, edited by Bird and Vaillancourt, World Bank.
--------------------. 2006. “Local and Regional Revenues: Realities and Prospects,” in
Perspectives on Fiscal Federalism, edited by Bird and Vaillancourt, World Bank.
Bird, Richard M. and Francois Vaillancourt. 2006. “Introduction and Summary,” in
Perspectives on Fiscal Federalism, ed. by Bird and Vaillancourt, World Bank.
Broadway, Robin, Sandra Roberts, and Anwar Shah. 1994. The Reform of Fiscal Systems
in Developing and Emerging Market Economics: A Federalism Perspective,
Policy Research Working Paper 1259, the World Bank, Washington, D.C.
Broadway, Robin and Anwar Shah. 2007. “Overview,” in Intergovernmental Fiscal
Transfers: Principles and Practice, ed. by Robin Broadway and Anwar Shah, the
World Bank.
Cai, Honbin and Daniel Treisman. forthcoming. “Did government decentralization cause
China’s economic miracle?” World Politics.
22
Cremer, Jacques, Antonio Estache, and Paul Seabright. 1994. The Decentralization of
Public Services: Lessons from the Theory of the Firm, Policy Research Working
paper 1345, World Bank, Washington, D.C.
Devarajan, Shanta, Stuti Khemani, and Shekhar Shah. 2007. The Politics of Partial
Decentralization, the World Bank.
Downs, Anthony. 1957. An Economic Theory of Democracy, Harper and Brothers.
Ebel, Robert D., Serdar Yilmaz. 2002. On the Measurement and Impact of Fiscal
Decentralization, World Bank.
---------------. 2002. Concept of Fiscal Decentralization and Worldwide Overview,
Working Paper No. 30346, World Bank.
Head, John, G. 1974. “Public Goods and Public Policy,” Public Good and Welfare, John
G. Head (ed.), Duke University Press.
Manor, James. 1999. The Political Economy of Democratic Decentralization, World
Bank, Washington, DC.
Martinez-Vazquez, Jorge, Charles McLure, and Francois Vaillancourt. 2006. “Revenues
and Expenditures in an Intergovernmental Framework,” in Perspectives on Fiscal
Federalism, edited by Bird and Vaillancourt, World Bank.
McLure, Charles E. Jr. 1995. “Comment on ‘Fiscal Federalism and Decentralization: A
Review of Some Efficiency and Macroeconomic Aspects,’ by Vito Tanzi,” Bruno,
Michael and Boris Pleskovic (editors), Annual Bank Conference on Development
Economics 1995, World Bank.
Musgrave, Richard. 1959. The Theory of Public Finance, (New York; McGraw-Hill).
Oates, Wallace E. 1972. Fiscal Federalism, New York: Harcourt, Brace, Jovanovich.
Olson, Mancur. 1969. “The Principle of ‘Fiscal Equivalence’: The Division of
Responsibilities Among Different Levels of Government,” American Economic
Review, vol. 59 (May).
Peters, Alan and Peter S. Fisher. 2002. State Enterprise Zone Programs, pp. 1-14, 217237.
Prud’homme. Remy. 1994, On the Dangers of Decentralization, Policy Working Paper
1252, World Bank, Washington, D.C.
---------------. 1996. “Comment on Conflicts and Dilemmas of Decentralization by Rudolf
Hommes,” Bruno, Michael and Boris Pleskovic (editors), Annual Bank
Conference on Development Economics 1995, World Bank.
Rawls, John. 1971. A Theory of Justice, The Belknap Press of Harvard University Press,
Cambridge, Massachusetts, London, England.
Rodden, Jonathan, Gunnar Eskeland, and Jennie Litvack. 2000. Fiscal Decentralization
and the Challenge of Hard Budget Constraints, World Bank.
Shah, Anwar, and Sana Shah. 2006. “The New Vision of Local Governance and the
Evolving Roles of Local Governments,” in Local Governance in Developing
Countries, edited by Anwar Shah, World Bank.
Sappington, David E. M. and Joseph E. Stiglitz. 1987. “Privatization, Information and
Incentives,” Journal of Policy Analysis and Management, Vol. 6, No. 4, 567-582.
Sen, Amartya K. 1970 (1984). Collective Choice and Social Welfare, Elsevier Science
Publishers.
23
--------------------. 2000. “Social Justice and Distribution of Income,” in Atkinson and
Bourguignon (editors), Handbook of Income Distribution, Elsevier Science
Publishers.
Stiglitz, Joseph. 1998. "The Private Uses of Public Interests: Incentives and Institutions,”
Journal of Economic Perspectives, Vol. 12, No. 2, Spring (1998), 3-22.
Tanzi, Vito. 1996. “Fiscal Federalism and Decentralization: A Review of Some
Efficiency and Macroeconomic Aspects,” Bruno, Michael and Boris Pleskovic
(editors), Annual Bank Conference on Development Economics 1995, World
Bank.
Tiebout, Chares. 1956. “A Pure Theory of Local Expenditure,” Journal of Political
Economy, vol. 64 (Oct.).
Tommasi, Mariano. 2006. “Federalism in Argentina and the Reforms of the 1990s,” in
Wallack and Srinivasan (editors), Federalism and Economic Reforms, Cambridge
University Press, New York, 2006.
White, Roland and Paul Smoke. 2005. “East Asia Decentralizes,” in East Asia
Decentralizes: Making Local Government Work, ed. Roland White, the World
Bank.
Wolman, Harold. 1994. “Decentralization: What Is It and Why We Should Care,” in
Robert J. Bennett (ed.), Decentralization, Local Governments, and Markets,
Clarendon Press, Oxford.
Wolman, Harold, et al. 1992. “National Urban Economic Development Policy,” Journal
of Urban Affairs, Vol. ¾.
World Bank. 1997. World Development Report (WDR) 1997 – the State in a Changing
World, World Bank
---------------. 1999. World Development Report (WDR) 1999/2000 – Entering the 21st
Century: the Changing Development Landscape, World Bank.
---------------. 2004. World Development Report (WDR) 2004: Making Services Work for
Poor People, World Bank.
---------------. 2004. Berlin Workshop Series 2004: Service Provision for the Poor—
Public and Private Sector Cooperation, World Bank.
---------------. 2005. Annual World Bank Conference on Development EconomicsEurope: Are We on Track to Achieve the Millennium Development Goals? World
Bank.
Yilmaz, Serdar, Jozsef Hegedus, and Michael E. Bell. 2003. “Implementing Fiscal
Decentralization Data Needs and Availability,” in Subnational Data
Requirements for Fiscal Decentralization: Case Studies from Central and Eastern
Europe, ed. By Serdar Yilmaz, Jozserf Hergedus and Michael E. Bell, the World
Bank.
24
Appendix
This efficiency argument can be represented graphically by figure 1, in which i and j are
the per capita quantities of two local goods to be provided. LM is a budget constraint
showing the quantities of i and j that can be produced. A and B are (families of)
indifference curves for local governments A and B respectively. In a centralized system,
the central government will settle on c, on indifference curves A’ and B’. A’ and B’ are
indifference curves inferior to indifference curves A and B. In a decentralized system,
local government A will settle at point a (producing ia and ja), on indifference curve A;
and local government B will settle at b, on difference curve B. Thus, decentralized
provision implies a welfare gain compared with central provision.
Figure 1. Welfare Benefits of Decentralization
j
A'
A
B
B'
L
a
ja
c
b
ia
M
i
Source: Remmy Prud’homme (1994), pp. 6.
25